First, though, we need to go back in time. Despite the intervention of two world war bombing campaigns, an underground tube network and the Great Fire, the layout of much of central London hasn't really changed since Georgian, or even medieval times. This means narrow streets, lots of buildings of historical importance and not much room for maneouvre.
The city's roads and layout have evolved over hundreds of years, but the demand for further change is happening much faster than that. Five years ago, London didn't have cycle superhighways. Two years ago London didn't have segregated cycleways. Yet barely a day goes past without a new demand that London's roads 'must change'. And change right now.
So what's to be done?
We could do what Amsterdam did in the post-war period and knock down a few buildings to widen the roads and turn public spaces into car parks. In fact, it's not unheard of in London either — Lower Thames Street and Regent Street were widened after demolishing the odd building. Let's face it though, it's likely to be costly and disruptive as well as upsetting those who appreciate heritage and architecture. It also won't necessarily solve the problem — the majority of recent fatalities have been on wider modern roads and junctions even where there is cycling provision.
Banning motor vehicles (apart from buses) is a popular option in some quarters, but there are a few practicalities to consider before we free up all that lovely road space and convert it into quaint cobbled streets (with smooth bits for the cyclists). TfL's annual report for 2012 suggests a net revenue of £139m from the congestion charge. Sure, it's a relatively small percentage of TfL's gross revenue but it's there nonetheless. And with only about 7% of commuting journeys into central London being done by private car, someone might want to think about what all those commercial vehicles are actually doing there and find alternative options for deliveries and construction traffic. See also: taxis.
Parts of the square mile are significantly worse than other areas of London — anyone who's tried to walk down Lombard Street in rush hour will find themselves hopping into the road on a regular basis. Bank junction is a cumbersome spaghetti of roads at which decades of fettling have led to no-one being favoured, not even motorists. The Bank Area strategy aims to pedestrianise Poultry and either Threadneedle Street or Cornhill to improve walking and cycling. Elsewhere, some schemes have attempted to separate pedestrians and traffic in two vertical layers, as we still see in the Barbican today and through the (now disappearing) Pedway. But while we're still keen on building upwards, we seem to have lost our vision for incorporating people into it.
Is this the way forward for improving London's roads and pavements? Pockets of pedestrianisation combined with re-purposing wider roads and footpaths to carve out decent spaces for pedestrians and cyclists too? It won't work everywhere but that's the problem faced by anyone trying to apply a one-size-fits-all solution — the diversity of what's there already. If it was that easy, someone would have done it before.
Cycling commissioner Andrew Gilligan points out that putting in 'panic measures' is just as bad as doing nothing at all, and makes some valuable points about the impact of banning HGVs and construction traffic:
What, I've asked many cycle activists, what is it you want us to do that we're not doing already? The usual answer is "do it quicker". But we can't simply slap in panic changes that might make cyclists safer at the expense of other people's safety. And if we're reducing roads' traffic capacity, as we are, we need changes over a far wider area to reduce vehicle numbers coming in to those roads.
The mayor's Roads Task Force wants to inject £30bn into London's roads to try and address the problems, but if we want the right solutions we may have to wait for them.
Here's an animation of what the future of urban transport might look like.
Photo by Another Partial Success in the Londonist Flickr pool.